Matching demographic for testing is not important?

July 24, 2008
Published by Krzysztof P.

demographic_chart When selecting the sample of respondents for the purpose of testing with users, often most of us is guided by the statistics and marketing research aimed at defining target users ages tested website.

It is a comfortable, natural and obvious that the process of selecting the testers corresponded to the guidelines, which often gets to the department or the wider strategy - marketing. But is this the correct approach? Do demographic criterion should be that which determines to a large extent the choice of candidates for testing purposes?

When I read an article describing an interview with Dana Chisnell (author of the book
Handbook of Usability Testing), I began to think seriously, and actually looks like approach to the selection of individuals to test for me and how does it most of us, as is to organize, conduct and analyze such studies.

Basically, just as you mentioned above, the selection of age (demographic) seems the most obvious and natural way to define groups of users, of which we selected respondents.

But what if we look at it in a slightly different way ...

This is how users behave on the page, to a large extent determine their habits, skills and knowledge possessed. These behaviors usually verifies and limited service interface, which may not include grated approach using their behavior, which in turn may in some way restrict the freedom of users in using the service.

However, what is most important is that if you're a 20 year old does not make you a super specialist in information and computer science. Also that you are 60-year old does not mean that you are helpless internet user. By analogy: if you are interested in a loan for a car does not mean that you're 26-30 years old and wealthy on average. You might as well be 50 years of age, have a lot of money but now you have a temporary liquidity problems and the car you need.

Relying solely on age as a determinant of suitability for testing, belonging to the target and determine interest in the test subject is very slippery. Of course - the probability of finding a person in a specific age group correlated with selected traits is greater, but that does not mean that is equal to 100%. It is important, what is the actual determination of the user (whether he is interested in issues), it is important to what his actual knowledge of a topic, it is important what are his real needs. Only then it is important whether it is 25 or 55 year old.

Thus, relying solely on age as a determinant index of suitability and selection for the test group is very risky. Why? Since the results we get from these tests can be unreliable (this of course is always a risk). So important is the selection of suitable persons for testing.

Dana Chisnell proposes that at the very beginning in a few words to determine user characteristics which interests us. Only to choose adequate characterization of such a person. And of course possible to use a variety of factors that describe the potential tester.

To test the so-called. Bug-these are not really of great importance, whom we select. To test użytecznościowych is one of the most important factors and should seriously think about who is responsible for our needs. Uncritical approach to the argument of market research will bring only exclusively astray. In a word: everywhere you need to think!